NexoFileCompanion

V0.1 · DRAFT FOR SIGN-OFF

NexoFileCompanion

Make your files understandable to AI.

$ pip install nexofc && nexo asset.jpg

THE PROBLEM

Files hold the data. They don't hold the context.

01 · formats

Embedded metadata is fragile, proprietary, and inconsistent across formats.

02 · ai

AI, agents, and RAG pipelines need explicit context to answer accurately.

03 · tooling

Without a shared structure, every tool reinvents its own metadata layer.

THE ANSWER

A companion, next to the asset

"The asset contains the data. NEXO gives it context."

Better context. Better AI answers.

asset.jpg
└── travels alongside
asset.jpg.nexo.md

AS SIMPLE AS THAT

One command generates a valid companion

The asset remains the asset. NEXO travels along it.

$ pip install nexofc
$ nexo asset.jpg
generates asset.jpg.nexo.md:
---
nexo: "0.1"
kind: image
file:
  name: asset.jpg
  size: 4839201
  mimeType: image/jpeg
  sha256: "a3f52c8d4e19b74a3d29d8657c78f22eab..."
---

PRINCIPLES

What guides the spec's design

01

Local-first — context lives next to the file

02

Open by design — no proprietary dependency

03

Human-readable — editable in plain text

04

Machine-understandable — parseable and validatable

05

AI-ready — reduces ambiguity for grounding

06

No invention — unknown stays unknown

07

Durable context — curated content is never destroyed

08

Replaceable tooling — no vendor lock-in

09

Optional ontology — useful even with zero enrichment

10

Outlives the app — survives the tool that created it

STATED LIMITS

What NEXO does not promise

✕

Does not promise universal gains in AI accuracy — only better-structured context.

✕

Does not verify truth — extracted means "as found", generated is labeled, not endorsed.

✕

Does not replace a database, ontology, or DAM/DMS.

✕

Does not require AI to be useful — a companion can be written entirely by hand.

✕

Does not alter, embed into, or execute anything from the original asset.

NEXO CORE 0.1

A small base, no dependencies

The Core defines only the companion's shape and the identity of the file it describes (nexo, kind, file.*). Zero dependency on AI, database, ontology, or external tool.

Vocabulary, provenance, and namespaces live outside the Core, as optional layers.

A Core 0.1 companion MUST:

✓be UTF-8 text
✓be Markdown
✓begin with valid YAML front matter
✓have a YAML mapping at the root
✓contain nexo at the root
✓contain file.name
✓contain file.sha256
---
nexo: "0.1"
file:
  name: photo.jpg
  sha256: "a3f52c8d4e19b74a3d29d865...46c32"
---

CORE FIELDS · nexo · kind · file

nexoSpecification version — stringrequired
kindGeneral type: image, audio, video, document, text, archive, dataset, otheroptional
file.nameCurrent file namerequired
file.sizeSize in bytesoptional
file.mimeTypeKnown MIME typeoptional
file.sha256SHA-256 hash, 64 hex charactersrequired
file.md5MD5 hash, 32 hex charactersoptional

TWO LAYERS

YAML for facts. Markdown for knowledge.

YAML holds facts that are short, structured, and searchable. Markdown holds long-form content: description, transcription, summary, notes, and human context.

--- yaml front matter ---
# Summary# Text Content# Transcription# Statistics# Notes

PROVENANCE

A value describes the asset. Provenance explains how it was obtained.

→Review and verification workflows
→Reproducibility of enrichment pipelines
→Detection of suspicious or conflicting metadata
→Preserves the line between source fact and interpretation

Five ways to say "where it came from"

declaredStated by a person or declarative source
extractedObtained directly from the asset or embedded metadata
inferredDeduced by a deterministic rule or classification
generatedProduced by an AI model or generative process
importedCopied from an external source or system

Validation status — tracked separately from the method

unverified — not reviewed against the source

reviewed — inspected by a person or process

verified — checked against a defined source or procedure

Example — suspicious metadata preserved, not promoted to fact:

image:
  exif:
    dateTime: "2099:11:24 23:46:47"
provenance:
  image.exif.dateTime:
    method: extracted
    validation: unverified

VOCABULARY 0.1

Standardized descriptive fields

title / displayTitledescriptionlanguagecreated / creatortagsoriginidentifiersrights

Echoes Dublin Core / schema.org where it makes sense — no new terminology invented without need.

TRUST SIGNAL

agentReady — computed, never declared

1. Structurally valid (Core)

2. sha256 verified against actual bytes

3. Audit found no errors or hygiene findings

Agent Ready — verified by nexo

Only shown when the technical field is true in that run of nexo audit — never a loose marketing claim.

NAMESPACES

Type-specific fields per asset kind

Technical container ≠ work/edition. Sibling namespaces carry editorial metadata next to the technical one.

image+ camera
Capture-device metadata when it exists
audio+ music · episode
Musical work or podcast/broadcast episode
video+ camera
Capture-device metadata when it exists
document+ book
When it is a book/publication
text+ code
When the text is source code
archive
dataset

kind: other is the catch-all with no matching namespace. companion.* holds metadata about the document itself — createdAt never changes; updatedAt is refreshed whenever the companion is written.

NEXO-CLI 0.1

Six commands, Core scope

nexo <asset>                  # create a companion
nexo inspect <asset>.nexo.md  # human-readable summary
nexo validate <asset>.nexo.md # structural conformance
nexo audit <asset>.nexo.md    # full conformance check
nexo collect <folder>         # summarize companions
nexo --scan <folder> --recursive

nexo audit — the full conformance and integrity check

1. structureEverything validate checks: parses, has nexo, file.name, and file.sha256
2. integrityRecomputes the asset's sha256 and compares against the declared value
3. hygieneScans for secrets, tokens, and hidden AI instructions

Key options — explicit control, nothing silent

--dry-runShows intended changes without writing
--forceDestructive overwrite, never the default
--offlineBlocks network access by providers
--outputWrites companion to a specific path
--minimalfile.* and kind only — no providers, no body
--dotfileWrites the hidden-name companion variant
--provider / --no-aiReserved for future enrichment phases

Exit codes — stable across 0.x

0success
1generic error
2usage error
3companion already exists
4validation failed
5hash mismatch
6asset unreadable/missing
7dependency failure
8hygiene finding

OVERFLOW FILES

A lean companion, even with large content

A 2-hour transcript or the OCR of a 300-page book can become a separate .nexo.<section>.md file, referenced by a standard Markdown link.

Overflow can be turned off and limits can be set.

meeting.mp3
meeting.mp3.nexo.md # Transcription See [meeting.mp3.nexo.transcript.md](…)
└── links to
meeting.mp3.nexo.transcript.md

REAL COMPANIONS

One spec, every kind of file

Seven real companions, one after another — each with the raw .nexo.md file shown in full, plus the rendered preview and downloads.

01
ogg · voice noteWhatsApp Ptt 2026-08-18 at 11.48.22.ogg
WhatsApp Ptt 2026-08-18 at 11.48.22.ogg.nexo.md
WhatsApp Ptt 2026-08-18 at 11.48.22.ogg.nexo.md.nexo.md
---
nexo: "0.1"
kind: audio

file:
  name: "(audio WhatsApp message) WhatsApp Ptt 2026-08-18 at 11.48.22.ogg"
  size: 137228
  mimeType: audio/ogg
  sha256: "ea1e1722c8fcdf131e2f9cfeaea23ba4867f6f7ddc18d1dbc0470889c4a3e5f0"
  md5: "38e5e7f80f138512aa19dfeabd4a7185"

language: en

tags:
  - WhatsApp
  - voice-message
  - family
  - invitation
  - lunch
  - paella

transcription:
  format: basic
  language: en
  provider: faster-whisper
  model: large-v3
  
whatsapp:
  messageType: voiceMessage
  direction: incoming
  sentAt: "2026-08-18T11:48:22"
  sender:
    name: "John Smith"
    phone: "+1 212 555-0147"
  chat:
    type: direct
    name: "Mary Smith"
  forwarded: false
  replyTo: null
  messageId: "3EB0F8A4D7C1E29B6F35"
  
provenance:
  file.name:
    method: declared
  file.size:
    method: extracted
  file.mimeType:
    method: extracted
  file.sha256:
    method: extracted
  file.md5:
    method: extracted
  kind:
    method: inferred
  language:
    method: inferred
  tags:
    method: generated
  transcription:
    method: generated
    validation: unverified
---

# Transcription

Hey, Mom. I was thinking about you today, and I realized just how much I miss you. It feels like we haven't had a proper chance to talk in a while, so I wanted to send you a quick message and let you know that everything's good over here. Work's been pretty busy, as usual, but things are going well. I've got a lot going on and a few new challenges, but nothing bad. Actually, I'm really enjoying it and learning a lot.

I also wanted to ask if you'd like to come over for lunch next Saturday. I'm making paella. I learned how to make it from a Spanish chef who showed me a few tricks to get the rice and seafood just right. So now I need someone I trust to try it and tell me if I actually learned anything, and who better than you?

Come over, take it easy, and spend the afternoon with us. We can have lunch, have a drink, chat, and properly catch up. I miss you, Mom. It would be really nice to spend some time together. I hope you can make it. Kisses!

RENDERED PREVIEW

▸

When the transcription is not large it goes into the .nexo.md file.

02
py · codenexo_orphan_scanner.py
nexo_orphan_scanner.py.nexo.md
nexo_orphan_scanner.py.nexo.md.nexo.md
---
nexo: '0.1'
kind: text
file:
  name: (code py) nexo_orphan_scanner.py
  size: 4183
  mimeType: text/x-python
  sha256: 4b9d3b457c3edd4c98bb9d10a092c25b6aed17c21eb3a31a033e949e187ea896
  md5: 368ce07bc5da80339a61687f83fe9aaa

text:
  encoding: utf-8
  lines: 126

code:
  language: python

tags:
- nexo
- nexofc
- cli-tool
- companion-audit
- python
- developer-tooling

provenance:
  kind:
    method: inferred
    validation: verified
  file.name:
    method: declared
    validation: reviewed
  file.size:
    method: extracted
    validation: verified
  file.mimeType:
    method: extracted
    validation: verified
  file.sha256:
    method: extracted
    validation: verified
  file.md5:
    method: extracted
    validation: verified
  text:
    method: extracted
    validation: verified
  code.language:
    method: inferred
    validation: verified
  title:
    method: declared
    validation: reviewed
  language:
    method: inferred
    validation: unverified
  tags:
    method: generated
    validation: unverified
  summary:
    method: generated
    validation: unverified
---

# Summary

`nexo_orphan_scanner.py` is a standalone Python 3 utility written for the nexoFC project. It is not part of the reference `nexo` CLI — it is an example of a small third-party tool built on top of the NEXO File Companion format, showing what an independent implementation can do with nothing more than the file-naming convention the format defines.

A NEXO companion always sits beside its asset as `<asset>.nexo.md`. Over time a folder tree can drift out of sync with that convention in two ways: an asset that was never processed and has no companion, or a companion left behind after its asset was renamed, moved, or deleted. This script walks a directory recursively, pairs every asset with a companion by file name, and reports both kinds of drift in a single run.

It also supports an optional `--verify-hashes` pass: for every correctly paired asset and companion, it recomputes the asset's SHA-256 and compares it against the `file.sha256` value declared in the companion's YAML front matter, flagging any mismatch. This mirrors, in miniature, the integrity check that `nexo audit` performs (`nexo-cli-0.1.md` §2.1) — recomputing a hash from the asset's actual bytes rather than trusting a stored value.

The script is deliberately read-only: it never writes, deletes, or modifies an asset or a companion, and it never invents a value it did not compute. It is meant as a reference example for anyone building tooling — indexers, backup auditors, digital-asset-management integrations — around the NEXO format rather than around the `nexo` CLI itself, reinforcing the Manifesto's point that NEXO content should stay useful without the application that produced it.

Usage:

```text
python nexo_orphan_scanner.py ./library --verify-hashes
```

RENDERED PREVIEW

▸

code is a sibling of text, when the text is source code.

03
mp3 · musicPachebell - Canon and Gigue in D major.mp3
Pachebell - Canon and Gigue in D major.mp3.nexo.md
Pachebell - Canon and Gigue in D major.mp3.nexo.md.nexo.md
---
nexo: "0.1"
kind: audio

file:
  name: "(audio music) Pachebell - Canon and Gigue in D major.mp3"
  size: 6287392
  mimeType: audio/mpeg
  sha256: "76b443c3f59fe33092e29379d3756edf936c4938716177869b953ae2eed01137"
  md5: "b436dfcd972a4fdf6a7aa2e1803894da"
  createdAt: "2026-08-17T10:37:47-03:00"
  modifiedAt: "2026-08-17T10:37:48-03:00"

companion:
  createdAt: "2026-08-17T10:38:15-03:00"
  updatedAt: "2026-08-17T10:38:15-03:00"

title: "Canon and Gigue in D Major for 3 violins and continuo"
description: "Recorded live in the Tapestry Room of the Isabella Stewart Gardner Museum, Boston."
identifiers:
  musicBrainzWork: "70ed831c-716e-3ba2-b997-527ea6ec1ca3"
rights:
  license: "CC BY-NC-ND 3.0"
language: zxx
tags:
  - Classical

audio:
  duration: 392.803313
  codec: mp3
  bitrate: 128000
  sampleRate: 44100
  channels: 2

music:
  title: "Canon and Gigue in D Major for 3 violins and continuo"
  artists:
    - "Chamber Music Society of Lincoln Center"
  album: "Music from the Isabella Stewart Gardner Museum, Boston"
  composers:
    - "Johann Pachelbel"
  track:
    number: 2
    total: 4
  work:
    title: "Canon and Gigue in D major, P. 37, T. 337"
    musicBrainzId: "70ed831c-716e-3ba2-b997-527ea6ec1ca3"
    catalogueNumbers:
      - "P. 37"
      - "T. 337"

provenance:
  kind:
    method: inferred
    validation: verified
  file.name:
    method: declared
    validation: reviewed
  file.size:
    method: extracted
    validation: verified
  file.mimeType:
    method: extracted
    validation: verified
  file.sha256:
    method: extracted
    validation: verified
  file.md5:
    method: extracted
    validation: verified
  file.createdAt:
    method: extracted
    validation: unverified
  file.modifiedAt:
    method: extracted
    validation: unverified
  companion:
    method: declared
  title:
    method: extracted
    validation: verified
  description:
    method: extracted
    validation: verified
  identifiers.musicBrainzWork:
    method: imported
    validation: verified
  rights.license:
    method: imported
    validation: reviewed
  language:
    method: inferred
    validation: reviewed
  tags:
    method: inferred
    validation: reviewed
  audio:
    method: extracted
    validation: verified
  music:
    method: extracted
    validation: verified
  music.work:
    method: imported
    validation: verified
---

RENDERED PREVIEW

▸

audio is regenerable via FFprobe. music comes from tags/MusicBrainz — editorial, not technical.

04
png · imageCorporateOffice.png
CorporateOffice.png.nexo.md
CorporateOffice.png.nexo.md.nexo.md
---
nexo: "0.1"
kind: image

file:
  name: "(image AI generated no text) CorporateOffice.png"
  mimeType: image/png
  size: 3516202
  sha256: "3704fb5398d1e1cae28a145af2bbf8c33b1fb60995a4ea86bc33c640db46bb64"
  md5: "e2052fc57ed6b022a818d48bacfb62c9"
  createdAt: "2026-08-15T19:35:43-03:00"
  modifiedAt: "2026-08-15T19:35:43-03:00"

companion:
  createdAt: "2026-08-15T19:36:10-03:00"
  updatedAt: "2026-08-17T17:26:13-03:00"

title: Corporate Office
description: "Modern multistory office building with glass façades and a stone central entrance, viewed from across a street with mature trees, palms, landscaped grounds, a driveway, and parked cars."
created: "2026-08-15T00:00:00Z"
tags:
  - office-building
  - corporate
  - commercial-architecture
  - glass-facade
  - building-exterior
  - trees
  - landscaped-grounds
  - parked-cars
  - daytime

image:
  format: png
  width: 1536
  height: 1024
  aspectRatio: "3:2"
  colorSpace: RGB
  channels: 3
  bitDepth: 8
  hasAlpha: false
  dpi: 72
  interlaced: false
  compressionMethod: deflate
  filterMethod: adaptive

provenance:
  kind:
    method: inferred
    validation: verified
  file.name:
    method: declared
    validation: reviewed
  file.size:
    method: extracted
    validation: verified
  file.mimeType:
    method: extracted
    validation: verified
  file.sha256:
    method: extracted
    validation: verified
  file.md5:
    method: extracted
    validation: verified
  file.createdAt:
    method: extracted
    validation: unverified
  file.modifiedAt:
    method: extracted
    validation: unverified
  companion:
    method: declared
  title:
    method: generated
    validation: unverified
  description:
    method: generated
    validation: unverified
  created:
    method: declared
    validation: unverified
  tags:
    method: generated
    validation: unverified
  image:
    method: extracted
    validation: verified
---

# Minhas notas (João)

Em julho do ano passado cheguei ao escritório no fim da manhã, sob um céu limpo e entre as árvores que emolduram a entrada principal. A reunião foi objetiva: conversamos sobre os próximos passos do projeto, as prioridades da equipe e a necessidade de transformar as decisões em ações simples e mensuráveis. Ao sair, fiquei com a impressão de que havia alinhamento suficiente para avançar — agora, o mais importante é manter o ritmo.
Nexo Example Note: "This section is in Portuguese because the person who wrote it typed in Portuguese. It is a free-form heading, not one of the recommended sections in `NEXO Core 0.1` §4 — NEXO does not restrict Markdown headings to that list."

RENDERED PREVIEW

Example 1 — office

Example 2 — flower

▸

image namespace plus camera metadata when capture-device data exists.

05
mp3 · interview1981 Interview Jacques Cousteau.mp3
+ overflow file
1981 Interview Jacques Cousteau.mp3.nexo.md

DOWNLOAD THE FILES

Original .mp3 is 133 MB — not hosted here. The two companion files above carry all its extracted context.

1981 Interview Jacques Cousteau.mp3.nexo.md.nexo.md
---
nexo: "0.1"
kind: audio

file:
  name: "(audio interview) 1981 Interview Jacques Cousteau.mp3"
  size: 133027242
  mimeType: audio/mpeg
  sha256: "fca07d58122de2861b713ef71b293c5c68a4d211a07bd30f61a8fca9b16a2d52"
  md5: "ec340e118ee060639fd73a83d7a941b8"

audio:
  duration: 3326
  codec: mp3
  bitrate: 320000
  sampleRate: 44100
  channels: 2

title: "Jacques Cousteau Discusses The Cousteau Almanac: An Inventory of Life on Our Water Planet"
creator:
  - type: person
    name: "Jacques-Yves Cousteau"
language: en
created: "1981-09-16"
description: "Captain Jacques-Yves Cousteau discusses The Cousteau Almanac: An Inventory of Life on Our Water Planet and the environmental movement in the United States and elsewhere."

tags:
  - interview
  - "spoken word"
  - oceanography
  - environment
  - ecology
  - environmental-protection
  - pollution
  - nuclear-energy
  - Cousteau-Society
  - United-Nations-Environment-Programme

episode:
  series:
    name: "Studs Terkel Radio Archive"
  publisher: "WFMT Radio Network"
  airDate: "1981-09-16"

transcription:
  format: basic
  language: en
  provider: faster-whisper
  runtime: local
  model: large-v3

provenance:
  file.name:
    method: declared
    validation: reviewed
  file.size:
    method: extracted
    validation: verified
  file.mimeType:
    method: extracted
    validation: verified
  file.sha256:
    method: extracted
    validation: verified
  file.md5:
    method: extracted
    validation: verified
  audio:
    method: extracted
    validation: verified
  title:
    method: extracted
    validation: verified
    source: id3
  creator:
    method: extracted
    validation: verified
    source: id3
  description:
    method: imported
    validation: reviewed
    source: "https://studsterkel.wfmt.com/programs/jacques-cousteau-discusses-his-book-cousteau-almanac-inventory-life-our-water-planet"
  tags:
    method: imported
    validation: reviewed
    source: "https://studsterkel.wfmt.com/programs/jacques-cousteau-discusses-his-book-cousteau-almanac-inventory-life-our-water-planet"
  episode:
    method: extracted
    validation: verified
    source: id3
  transcription:
    method: generated
    validation: unverified
    provider: faster-whisper
    runtime: local
    model: large-v3
---

# Transcription

See [(audio interview) 1981 Interview Jacques Cousteau.mp3.nexo.text.md](%28audio%20interview%29%201981%20Interview%20Jacques%20Cousteau.mp3.nexo.text.md).

RENDERED PREVIEW

▸

Once the transcription is too large it overflows to a separate file, easily accessed through a standard Markdown link.

▸

Transcription is one possibility. Profiles can also be created for high-value analyses such as speaker identification, sentiment and emotion analysis, topic extraction, named entity recognition, concise summarization, language detection, keyword extraction and content moderation. NEXO is flexible and creativity is the limit.

OVERFLOW FILE1981 Interview Jacques Cousteau.mp3.nexo.text.md
▸

This transcript goes on for +1600 lines, so it is kept separate from the .nexo.md file, keeping it legible.

▸

Overflow can be turned off and limits can be set.

1981 Interview Jacques Cousteau.mp3.nexo.text.md.nexo.text.md
# Transcription

Source audio: [1981 Interview Jacques Cousteau - tagged.mp3](./1981%20Interview%20Jacques%20Cousteau%20-%20tagged.mp3)

Language: English

When we were young in school we were taught in history books about explorers.
The name Magellan, Balboa, Admiral Perry,
Amundsen, well it's a new kind of explorer today.
One who explores this organism called our environment.
One who explores deep down to the seas and sees the possibilities as well as
perils of its being abused. I think of course of Captain
Jacques-Yves Cousteau, the most celebrated of our explorers.
Of course, all of us are acquainted with him through his
quite marvelous television series.
And he and his young colleagues have put together a book.
It's called The Cousteau Ormanac and the Sub-Title Inventory
of Life on our Water Planet.
And it's beautiful.
It's thrilling.
But also, it's full of advice how to save this
environment from those who were the predators.
And it's about ordinary people who seem to know what it's about,
though they're not given that credit for that.
And so this is the book published by Doubleday Dolphin.
We'll talk about the Dolphin in a moment, too.
So Captain Cousteau, my guest, his thoughts
about the Ormanack, about us and where we live,
and how we live, and how we can live.
[MUSIC PLAYING]
Sailing down my golden river, sun and water all my own, yet I was never alone.
♪ Run man water, oh life givers ♪
♪ I'll have them wherever I roam ♪
♪ And I was not far from home ♪
- Just hearing those two verses of Pete Seeger's song
about the Hudson River where he lives
and that both the clear water.
Your thoughts, Captain Christof.
Well, Pete recorded for us a song along the Hudson River
for one of our films.
… 1585 more lines — download the full file below
06
epub · bookBaum, Frank - The Wonderful Wizard Of Oz.epub
+ overflow file
Baum, Frank - The Wonderful Wizard Of Oz.epub.nexo.md
Baum, Frank - The Wonderful Wizard Of Oz.epub.nexo.md.nexo.md
---
nexo: "0.1"
kind: document

file:
  name: "(document epub) Baum, Frank - The Wonderful  Wizard Of Oz.epub"
  size: 681182
  mimeType: application/epub+zip
  sha256: "be8f7802b737cf2f402960d4c691238ccd51b67de0cd69b909e35a2108da47a3"
  md5: "5fd6c612ebeea365006839d60bea121a"
  createdAt: "2026-08-17T10:24:26-03:00"
  modifiedAt: "2026-08-17T10:24:26-03:00"

companion:
  createdAt: "2026-08-17T10:25:03-03:00"
  updatedAt: "2026-08-17T10:25:03-03:00"

title: "The Wonderful Wizard of Oz"
creator:
  - type: person
    name: "L. Frank Baum"
language: en-US
description: "The adventures of a young farm girl named Dorothy in the magical Land of Oz, after she and her pet dog Toto are swept away from their Kansas home by a cyclone."
identifiers:
  projectGutenberg: "https://www.gutenberg.org/ebooks/55"
  internetArchive: "https://archive.org/details/wonderfulwizardo00baumiala"
origin:
  url: "https://standardebooks.org/ebooks/l-frank-baum/the-wonderful-wizard-of-oz"
  retrievedDate: "2026-08-17"
tags:
  - "Oz (Imaginary place) -- Juvenile fiction"
  - "Fantasy literature"
  - "Courage -- Juvenile fiction"
  - "Home -- Juvenile fiction"
  - "Good and evil -- Juvenile fiction"
  - "Dreams -- Juvenile fiction"
  - "Gale, Dorothy (Fictitious character) -- Juvenile fiction"
  - "Cowardly Lion (Fictitious character) -- Juvenile fiction"
  - "Cyclones -- Juvenile fiction"
  - "Scarecrow (Fictitious character from Baum) -- Juvenile fiction"
  - "Toto (Fictitious character) -- Juvenile fiction"
  - "Witches -- Juvenile fiction"
  - "Wizard of Oz (Fictitious character) -- Juvenile fiction"
  - "Tin Woodman (Fictitious character) -- Juvenile fiction"
  - Fiction
rights: >-
  The source text and artwork in this ebook are believed to be in the United
  States public domain; that is, they are believed to be free of copyright
  restrictions in the United States. They may still be copyrighted in other
  countries, so users located outside of the United States must check their
  local laws before using this ebook. The creators of, and contributors to,
  this ebook dedicate their contributions to the worldwide public domain via
  the terms in the [CC0 1.0 Universal Public Domain
  Dedication](https://creativecommons.org/publicdomain/zero/1.0/).

book:
  title: "The Wonderful Wizard of Oz"
  publisher: "Standard Ebooks"
  publicationDate: "2026-08-15"
  series:
    name: Oz
    position: 1

provenance:
  kind:
    method: inferred
    validation: verified
  file.name:
    method: declared
    validation: reviewed
  file.size:
    method: extracted
    validation: verified
  file.mimeType:
    method: extracted
    validation: verified
  file.sha256:
    method: extracted
    validation: verified
  file.md5:
    method: extracted
    validation: verified
  file.createdAt:
    method: extracted
    validation: unverified
  file.modifiedAt:
    method: extracted
    validation: unverified
  companion:
    method: declared
  title:
    method: extracted
    validation: verified
  creator:
    method: extracted
    validation: verified
  language:
    method: extracted
    validation: verified
  description:
    method: extracted
    validation: verified
  identifiers:
    method: extracted
    validation: verified
  origin.url:
    method: declared
    validation: reviewed
  origin.retrievedDate:
    method: inferred
    validation: unverified
  tags:
    method: extracted
    validation: verified
  rights:
    method: extracted
    validation: verified
  book:
    method: extracted
    validation: verified
  textContent:
    method: extracted
    validation: verified
---

# Text Content

See [(document epub) Baum, Frank - The Wonderful  Wizard Of Oz.epub.nexo.text.md](%28document%20epub%29%20Baum%2C%20Frank%20-%20The%20Wonderful%20%20Wizard%20Of%20Oz.epub.nexo.text.md).

RENDERED PREVIEW

▸

Overflow — same case as transcription: the full text goes to its own file.

OVERFLOW FILEBaum, Frank - The Wonderful Wizard Of Oz.epub.nexo.text.md
▸

The text continues for +2300 lines.

Baum, Frank - The Wonderful Wizard Of Oz.epub.nexo.text.md.nexo.text.md
Yet the old time fairy tale, having served for generations, may now be classed as “historical” in the children’s library; for the time has come for a series of newer “wonder tales” in which the stereotyped genie, dwarf and fairy are eliminated, together with all the horrible and bloodcurdling incidents devised by their authors to point a fearsome moral to each tale. Modern education includes morality; therefore the modern child seeks only entertainment in its wonder tales and gladly dispenses with all disagreeable incident.

Having this thought in mind, the story of The Wonderful Wizard of Oz was written solely to please children of today. It aspires to being a modernized fairy tale, in which the wonderment and joy are retained and the heartaches and nightmares are left out.

L. Frank Baum

Chicago, April, 1900.

## The Wonderful Wizard of Oz

## I

The Cyclone

Dorothy lived in the midst of the great Kansas prairies, with Uncle Henry, who was a farmer, and Aunt Em, who was the farmer’s wife. Their house was small, for the lumber to build it had to be carried by wagon many miles. There were four walls, a floor and a roof, which made one room; and this room contained a rusty looking cookstove, a cupboard for the dishes, a table, three or four chairs, and the beds. Uncle Henry and Aunt Em had a big bed in one corner, and Dorothy a little bed in another corner. There was no garret at all, and no cellar—except a small hole dug in the ground, called a cyclone cellar, where the family could go in case one of those great whirlwinds arose, mighty enough to crush any building in its path. It was reached by a trap door in the middle of the floor, from which a ladder led down into the small, dark hole.

When Dorothy stood in the doorway and looked around, she could see nothing but the great gray prairie on every side. Not a tree nor a house broke the broad sweep of flat country that reached to the edge of the sky in all directions. The sun had baked the plowed land into a gray mass, with little cracks running through it. Even the grass was not green, for the sun had burned the tops of the long blades until they were the same gray color to be seen everywhere. Once the house had been painted, but the sun blistered the paint and the rains washed it away, and now the house was as dull and gray as everything else.

When Aunt Em came there to live she was a young, pretty wife. The sun and wind had changed her, too. They had taken the sparkle from her eyes and left them a sober gray; they had taken the red from her cheeks and lips, and they were gray also. She was thin and gaunt, and never smiled now. When Dorothy, who was an orphan, first came to her, Aunt Em had been so startled by the child’s laughter that she would scream and press her hand upon her heart whenever Dorothy’s merry voice reached her ears; and she still looked at the little girl with wonder that she could find anything to laugh at.

Uncle Henry never laughed. He worked hard from morning till night and did not know what joy was. He was gray also, from his long beard to his rough boots, and he looked stern and solemn, and rarely spoke.

It was Toto that made Dorothy laugh, and saved her from growing as gray as her other surroundings. Toto was not gray; he was a little black dog, with long silky hair and small black eyes that twinkled merrily on either side of his funny, wee nose. Toto played all day long, and Dorothy played with him, and loved him dearly.

Today, however, they were not playing. Uncle Henry sat upon the doorstep and looked anxiously at the sky, which was even grayer than usual. Dorothy stood in the door with Toto in her arms, and looked at the sky too. Aunt Em was washing the dishes.

From the far north they heard a low wail of the wind, and Uncle Henry and Dorothy could see where the long grass bowed in waves before the coming storm. There now came a sharp whistling in the air from the south, and as they turned their eyes that way they saw ripples in the grass coming from that direction also.

Suddenly Uncle Henry stood up.

“There’s a cyclone coming, Em,” he called to his wife. “I’ll go look after the stock.” Then he ran toward the sheds where the cows and horses were kept.

Aunt Em dropped her work and came to the door. One glance told her of the danger close at hand.

“Quick, Dorothy!” she screamed. “Run for the cellar!”

Toto jumped out of Dorothy’s arms and hid under the bed, and the girl started to get him. Aunt Em, badly frightened, threw open the trap door in the floor and climbed down the ladder into the small, dark hole. Dorothy caught Toto at last and started to follow her aunt. When she was halfway across the room there came a great shriek from the wind, and the house shook so hard that she lost her footing and sat down suddenly upon the floor.

Then a strange thing happened.

… 2295 more lines — download the full file below
07
pdf · report2026-08WorldDevelopmentReport2026ThePromiseofArtificialIntelligence.pdf
+ overflow file
…ArtificialIntelligence.pdf.nexo.md
…ArtificialIntelligence.pdf.nexo.md.nexo.md
---
nexo: "0.1"

kind: document

file:
  name: "(document pdf text) 2026-08WorldDevelopmentReport2026ThePromiseofArtificialIntelligence.pdf"
  mimeType: application/pdf
  size: 12600418
  sha256: "f3d90a98d139be00e0635f88b4771d32da0a807682cead11718f61cf42906dd5"
  md5: "0d8436256ddf274cbca19159d9921828"
  createdAt: "2026-08-15T15:39:51-03:00"
  modifiedAt: "2026-08-15T15:39:54-03:00"

companion:
  createdAt: "2026-08-15T15:40:20-03:00"
  updatedAt: "2026-08-15T15:40:20-03:00"

document:
  format: pdf
  pages: 380

title: "World Development Report 2026: The Promise of Artificial Intelligence"

creator:
  - type: organization
    name: World Bank

language: en
created: "2026"

identifiers:
  isbn: "978-1-4648-2331-2"
  eisbn: "978-1-4648-2357-2"
  doi: "10.1596/978-1-4648-2331-2"

rights:
  license: "CC BY 3.0 IGO"

description: "World Bank flagship report examining the opportunities, risks, adoption, adaptation, governance and development impacts of artificial intelligence, with particular attention to developing economies."

tags:
  - artificial-intelligence
  - development
  - public-policy
  - governance
  - productivity
  - developing-economies

provenance:

    method: declared
    validation: reviewed
  file.size:
    method: extracted
    validation: verified
  file.mimeType:
    method: extracted
    validation: verified
  file.sha256:
    method: extracted
    validation: verified
  file.md5:
    method: extracted
    validation: verified
  file.createdAt:
    method: extracted
    validation: unverified
  file.modifiedAt:
    method: extracted
    validation: unverified
  companion:
    method: declared
  document:
    method: extracted
  title:
    method: extracted
  creator:
    method: extracted
  language:
    method: extracted
  created:
    method: extracted
  identifiers:
    method: extracted
  rights:
    method: extracted
  description:
    method: generated
  tags:
    method: generated
  textContent:
    method: extracted
    validation: verified
---

# Minhas notas (João)

Para o Brasil, o relatório indica que a prioridade deve ser adaptar IA a problemas públicos e produtivos concretos — especialmente educação, gestão administrativa, fiscalização e apoio a pequenas empresas — em vez de concentrar recursos na criação de modelos de fronteira; o país já possui um ativo relevante de interoperabilidade no Pix, que pode inspirar uma infraestrutura digital aberta, competitiva e menos dependente de um único fornecedor. Os exemplos brasileiros citados mostram potencial em planejamento pedagógico, correção de textos, análise jurídica e previsão de riscos de corrupção, mas o ganho dependerá de dados públicos confiáveis, integração aos fluxos de trabalho, capacitação de servidores e professores, avaliação contínua, proteção contra vieses e contratos que preservem controle público e portabilidade.
Nexo Example Note: "This section is in Portuguese because the person who wrote it typed in Portuguese. It is a free-form heading, not one of the recommended sections in `NEXO Core 0.1` §4 — NEXO does not restrict Markdown headings to that list."

# Description

The report examines artificial intelligence as a general-purpose technology and analyzes its implications for low- and middle-income countries.

It organizes the discussion around AI capabilities, concentration and complementary conditions, and evaluates economic, social and political impacts.

# Summary

World Development Report 2026 presents artificial intelligence as a fast-diffusing general-purpose technology that could help address persistent development constraints for the 6.8 billion people living in low- and middle-income countries. Its central framework is that AI's development impact depends on three conditions: its capabilities to support complex cognitive tasks, the concentration of models, chips, and data centers in a small number of firms and economies, and the availability of complementary infrastructure, skills, data, and institutions.

The report recommends that most developing countries adopt existing tools, adapt them to local languages, data, and needs, and advance frontier AI only where resources and readiness justify it. It assesses possible gains in productivity, public services, and accountability alongside risks from unequal access, labor disruption, bias, misinformation, cybercrime, and dependence on foreign providers. It calls for sustained investment in electricity, connectivity, education, computing access, local data, institutional capacity, evidence, and public trust; governments should select problems carefully, evaluate systems continuously, and procure AI in ways that preserve public control and avoid vendor lock-in.


# Text Content

See [(document pdf text) 2026-08WorldDevelopmentReport2026ThePromiseofArtificialIntelligence.pdf.nexo.text.md](%28document%20pdf%20text%29%202026-08WorldDevelopmentReport2026ThePromiseofArtificialIntelligence.pdf.nexo.text.md).

RENDERED PREVIEW

▸

Overflow — same case as transcription: the extracted text goes to its own file.

OVERFLOW FILE…ArtificialIntelligence.pdf.nexo.text.md
▸

Original PDF: 12.6MB. NEXO files combined (.nexo.md + .nexo.text.md): 1.3MB — size is now 10,3%, content remains 100%.

▸

For assets from authoritative public sources, the original file may even be discarded when long-term preservation of its exact representation is not required.

…ArtificialIntelligence.pdf.nexo.text.md.nexo.text.md
# Text Content

The Promise of Artificial Intelligence

## The Promise of Artificial Intelligence
This book, along with any associated content or subsequent updates, can be accessed at https://hdl.handle.net/10986/45296.
## AI Disclosure Statement
Claude Sonnet 4.6 and 5, DeepSeek Chat, GPT-5 mini, Gemini 1.5 Flash, and Opus 4.6 and 4.8 were used for coding and labeling unstructured data. The World Bank's internal mAI interface using ChatGPT 5 and Copilot was used to improve clarity and shorten text in earlier drafts. All uses and outputs of AI were closely monitored by the team. All analytical judgments, policy interpretations, and con­ clusions were developed, reviewed, and verified by the author team.

Scan to see this and prior editions of the World Development Report .
## The Promise of Artificial Intelligence

© 2026 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org
## Some rights reserved
## 1 2 3 4   29 28 27 26
This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent.
The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. The boundaries, colors, denominations, links/footnotes, and other information shown in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. The citation of works authored by others does not mean The World Bank endorses the views expressed by those authors or the content of their works. Certain artificial intelligence ('AI') tools may have been used in creating this work, as indicated in the work, but all transformative and creative effort expressed in the work was performed by the World Bank staff or external contributors. The World Bank is not responsible for any claims, including for errors or inaccurate information, generated by AI programs not controlled or owned by the World Bank.
Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved.
## Rights and Permissions

This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http://creativecommons.org /­ licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions:
Attribution -Please cite the work as follows: World Bank. 2026. World Development Report 2026: The Promise of Artificial Intelligence . World Bank. doi:10.1596/978-1-4648-2331-2. License: Creative Commons Attribution CC BY 3.0 IGO
Translations -If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation .
Adaptations -If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank.
Third-party content -The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images.
All queries on rights and licenses should be addressed to World Bank Publications, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; e-mail: [email protected].
ISSN: 0163-5085
ISBN: 978-1-4648-2331-2
e-ISBN: 978-1-4648-2357-2
DOI: 10.1596/978-1-4648-2331-2
Artificial Intelligence: Page ii has information on the use of AI tools in the development of this work.
Cover image : The cover image was generated using OpenAI / ChatGPT image generation on June 24, 2026.
Cover design: Ivanna Locmanidis, Sergio Andres Moreno Tellez, and Greg Wlosinski, World Bank Creative Services.
© World Bank. Further permission required for reuse.
## The Library of Congress Control Number has been requested.
The manufacturer's authorized representative in the European Union for product safety is Mare Nostrum Group B.V., Doelen 72, 4831 GR Breda, The Netherlands. Email: [email protected].
## Contents
| xvii   | Foreword                                                                                                                                  |
|--------|-------------------------------------------------------------------------------------------------------------------------------------------|
| xxi    | Acknowledgments                                                                                                                           |
… 4499 more lines — download the full file below

LIGHTER BY DESIGN

Size is now 10,3%. Content remains 100%.

12.6MBoriginal pdf file
1.3MBnexo files combined

The original PDF carries the presentation layer — fonts and resources to reproduce it visually. NEXO files preserve factual metadata and textual content in a compact, open, UTF-8 format, cutting the cost of storage, indexing, retrieval, and AI processing.

World Development Report 2026 cover

NEXT STEPS

Help shape the standard from the start

→ Join https://github.com/nexo-fc. Nexo-cli will ship soon!

→ Feedback us at [email protected]

→ Review and comment on the specification (Core, Vocabulary, Provenance)

github.com/nexo-fc →
$nexo --scan . --recursive